Format: ORAL

Qin Li1, Kaining Hu2, Xiaojun Kou3, Richard Ree4

1 School of Ecological and Environmental Sciences, East China Normal University, Shanghai, China, 2 Department of Human Genetics, University of Chicago, Chicago, USA, 3 School of Life Sciences, Beijing Normal University, Beijing, Shanghai 4 Department of Science and Education, Field Museum of Natural History, Chicago, USA

High-altitude plant species have shown local adaptations to specific environments, such as low temperature, drought, ultraviolet radiation, and short growing season. In the face of climate change, alpine specialists are more likely to experience range contractions and evolutionary risks. To quantify such adaptation as well as vulnerability, we sampled 15 populations of an alpine Rhododendron (R. intricatum) with a large elevational range (3000 m - 4500 m) in the Mt. Gongga region of the eastern Hengduan Mountains in southwestern China. Using a whole-genome re-sequencing dataset, we performed genotype-environment association (GEA) and gene ontology (GO) enrichment analyses to explore its adaptation signature along elevation. Using inferred outlier SNPs (single nucleotide polymorphisms), we then modeled genetic turnover in a multidimensional climate space and estimated genetic offsets under climate change. The genomic data showed no significant geographic structure, while GAE tests revealed a large number of outlier SNPs associated with elevation. Among the inferred significant SNPs, GO enrichment analysis revealed a strong functional tendency associated with reproductive processes such as pollination and signaling. The current genetic turnover shows strong signals along climatic variables related to precipitation seasonality and extremes, which is consistent with a northeast-southwest geographic separation of the population. Meanwhile, such turnover signals would lead to remarkable risks at high elevations with intense precipitation changes in the future. Together, we highlight the genomic mechanisms underlying local adaptation at high altitudes. More importantly, we show that predicting climate vulnerability from both evolutionary and ecological factors reveals population-level risks to climate change in a mountainous region.